SciCode (AA)

Artificial Analysis' own SciCode run. Shown for reference.

Published by

Artificial Analysis

Category

Coding

Index weight

Reference only

Models

162

Data as of

9 Sept 2026

Source: Artificial Analysis (artificialanalysis.ai). Data taken from the public leaderboard.

1. 1Claude Fable 5.163.1%
2. 2Claude Fable 561.0%
3. 3Claude Fable 5.160.9%
4. 4Gemini 3.7 Flash59.8%
5. 5Muse Spark 1.359.7%
6. 6Kimi K359.5%
7. 7GLM-5.359.0%
8. 8Muse Spark 1.158.8%
9. 9Muse Spark 1.358.8%
10. 10Claude Fable 5.158.7%
11. 11Gemini 3.1 Pro58.7%
12. 12GPT-5.6 Sol57.8%
13. 13GPT-5.6 Sol57.4%
14. 14Muse Spark 1.257.4%
15. 15Gemini 3.7 Flash57.2%

SearchAll providersAI9StarsAlibabaAnt GroupAnthropicApodexArcee AICelerisCohereDeepSeekGoogleIBMInceptionInclusionAILG AI ResearchLiquid AILongCatMBZUAI Institute of Foundation ModelsMetaMiniMaxMistral AIMoonshot AIMotif TechnologiesMultiverse ComputingNex AGINVIDIAOpenAIOpenBMBSapiens AISK TelecomStepFunThinking MachinesUpstagexAIXiaomiZhipu AIOpen weights162 of 162

#	Model	Score ↓	Run ↕	Index ↕	Published ↕
1	Claude Fable 5.1Anthropic	63.1%	Claude Fable 5.1 (max with fallback)	70.0	–
2	Claude Fable 5Anthropic	61.0%	Claude Fable 5 (with fallback)	70.4	–
3	Claude Fable 5.1Anthropic	60.9%	–	70.3	–
4	Gemini 3.7 FlashGoogle	59.8%	–	64.4	–
5	Muse Spark 1.3Meta	59.7%	–	63.9	–
6	Kimi K3Moonshot AIopen	59.5%	Kimi K3 (max)	66.5	–
7	GLM-5.3Zhipu AIopen	59.0%	GLM-5.3 (max)	–	–
8	Muse Spark 1.1Meta	58.8%	Muse Spark 1.1 (xhigh)	64.9	–
9	Muse Spark 1.3Meta	58.8%	Muse Spark 1.3 (max)	–	–
10	Claude Fable 5.1Anthropic	58.7%	–	69.7	–
11	Gemini 3.1 ProGoogle	58.7%	–	63.9	–
12	GPT-5.6 SolOpenAI	57.8%	–	68.0	–
13	GPT-5.6 SolOpenAI	57.4%	–	66.0	–
14	Muse Spark 1.2Meta	57.4%	Muse Spark 1.2 (xhigh)	62.6	–
15	Gemini 3.7 FlashGoogle	57.2%	Gemini 3.7 Flash (high)	61.7	–
16	GPT-5.6 SolOpenAI	57.1%	–	67.5	–
17	GPT-5.6 SolOpenAI	57.1%	GPT-5.6 Sol (max)	65.3	–
18	Claude Fable 5.1Anthropic	56.7%	–	65.9	–
19	Gemini 3.8 FlashGoogle	56.6%	Gemini 3.8 Flash (high)	62.4	–
20	GPT-6 AstraOpenAI	56.5%	GPT-6 Astra (max)	69.9	–
21	Grok 4.6xAI	56.5%	Grok 4.6 (high)	63.6	–
22	Claude Opus 5Anthropic	56.4%	Claude Opus 5 (max)	70.0	–
23	Claude Fable 5.1Anthropic	56.4%	–	68.0	–
24	GPT-5.6 SolOpenAI	56.4%	–	64.6	–
25	GPT-5.5OpenAI	56.1%	–	66.3	–
26	Grok 4.6xAI	55.9%	–	63.8	–
27	GPT-5.5OpenAI	55.8%	GPT-5.5 (xhigh)	66.6	–
28	GPT-6 AstraOpenAI	55.7%	–	68.6	–
29	Claude Opus 5Anthropic	55.7%	–	66.3	–
30	Gemini 3.7 FlashGoogle	55.7%	–	62.5	–
31	GPT-6 AstraOpenAI	55.4%	–	69.5	–
32	Claude Opus 5Anthropic	55.4%	–	69.3	–
33	Gemini 3.8 FlashGoogle	55.1%	–	63.5	–
34	Grok 4.5xAI	55.0%	Grok 4.5 (high)	63.4	–
35	GPT-5.6 TerraOpenAI	55.0%	GPT-5.6 Terra (max)	62.3	–
36	Gemini 3.8 FlashGoogle	55.0%	–	61.9	–
37	GPT-5.5OpenAI	54.5%	–	64.9	–
38	Claude Opus 4.8Anthropic	54.4%	Claude Opus 4.8 (max)	64.7	–
39	Claude Sonnet 5Anthropic	54.3%	Claude Sonnet 5 (max)	61.2	–
40	Claude Sonnet 5Anthropic	54.3%	–	60.5	–
41	GPT-6 AstraOpenAI	54.2%	–	67.4	–
42	GPT-6 AstraOpenAI	54.0%	–	65.8	–
43	Claude Sonnet 5Anthropic	54.0%	–	57.3	–
44	Qwen3.8 2.4T A95BAlibabaopen	54.0%	–	–	–
45	Gemini 3.5 FlashGoogle	53.9%	Gemini 3.5 Flash	60.7	–
46	GPT-5.6 LunaOpenAI	53.6%	GPT-5.6 Luna (max)	56.6	–
47	GPT-6 AstraOpenAI	53.5%	–	64.4	–
48	Gemini 3.6 FlashGoogle	53.4%	Gemini 3.6 Flash	60.7	–
49	Qwen3 8Alibaba	53.2%	–	63.9	–
50	Grok 4.6xAI	53.0%	–	62.3	–
51	Kimi K3Moonshot AIopen	52.7%	–	58.8	–
52	GPT-5.5 InstantOpenAI	52.5%	–	60.4	–
53	GPT-5.6 TerraOpenAI	52.4%	–	63.1	–
54	GPT-5.6 TerraOpenAI	52.3%	–	64.2	–
55	GPT-5.4 miniOpenAI	52.1%	GPT-5.4 mini (xhigh)	54.8	–
56	GLM-5.3-FlashZhipu AIopen	51.6%	–	61.0	–
57	GPT-5.6 LunaOpenAI	51.6%	–	58.1	–
58	Claude Sonnet 5Anthropic	51.6%	–	–	–
59	Claude Opus 5Anthropic	51.5%	–	63.1	–
60	Kimi K2.6Moonshot AIopen	51.5%	Kimi K2.6	60.6	–

Show 100 more